EWS in measles data
2019-01-22
I use the spaero::get_stats() function to calculate EWS over a moving window for the measles case incidence data from four cities in Niger. I use a bandwidth of 35 weeks, resulting in a 70 week window. All EWS are calculated with backward_only = TRUE so that EWS are only calculated based on data in the past.
# Load packages -----------------------------------------------------------
library(tidyverse)
library(spaero)
# Load data ---------------------------------------------------------------
fname <- "../../data/clean-data/weekly-measles-incidence-niger-cities-clean.RDS"
measles_data <- readRDS(fname) %>%
filter(year > 1994) # drop first NA year, only used for modeling
# Calculate EWS -----------------------------------------------------------
all_stats <- tibble() # empty tibble for storage
for(do_region in unique(measles_data$region)){
cases <- measles_data %>%
filter(region == do_region) %>%
pull(cases)
city_stats <- spaero::get_stats(
x = cases,
center_trend = "local_constant",
center_kernel = "uniform",
center_bandwidth = 35,
stat_trend = "local_constant",
stat_kernel = "uniform",
stat_bandwidth = 35,
lag = 1,
backward_only = TRUE
)$stats
city_stats_tb <- as_tibble(city_stats) %>%
mutate(
time_iter = 1:n(),
date = unique(measles_data$date)
) %>%
gather(key = ews, value = value, -time_iter, -date) %>%
mutate(region = do_region)
all_stats <- bind_rows(all_stats, city_stats_tb)
}Define a function to plot the EWS and the observations.
plot_it <- function(dates, cases, ews, lab, title){
par(mar = c(5,5,2,5))
plot(dates, cases, type = "l", main = title)
par(new = T)
plot(dates, ews, type = "l", col = "red", axes=F, xlab=NA, ylab=NA)
axis(side = 4)
mtext(side = 4, line = 3, lab, cex = 0.7)
}Agadez
do_region <- "Agadez (City)"
dates <- unique(measles_data$date)
par(mfrow = c(4,3))
for(do_ews in unique(all_stats$ews)){
cases <- filter(measles_data, region == do_region) %>% pull(cases)
ews <- filter(all_stats, region == do_region & ews == do_ews) %>% pull(value)
plot_it(dates, cases, ews, lab = "ews", title = do_ews)
}
Maradi
do_region <- "Maradi (City)"
dates <- unique(measles_data$date)
par(mfrow = c(4,3))
for(do_ews in unique(all_stats$ews)){
cases <- filter(measles_data, region == do_region) %>% pull(cases)
ews <- filter(all_stats, region == do_region & ews == do_ews) %>% pull(value)
plot_it(dates, cases, ews, lab = "ews", title = do_ews)
}
Niamey
do_region <- "Niamey (City)"
dates <- unique(measles_data$date)
par(mfrow = c(4,3))
for(do_ews in unique(all_stats$ews)){
cases <- filter(measles_data, region == do_region) %>% pull(cases)
ews <- filter(all_stats, region == do_region & ews == do_ews) %>% pull(value)
plot_it(dates, cases, ews, lab = "ews", title = do_ews)
}
Zinder
do_region <- "Zinder (City)"
dates <- unique(measles_data$date)
par(mfrow = c(4,3))
for(do_ews in unique(all_stats$ews)){
cases <- filter(measles_data, region == do_region) %>% pull(cases)
ews <- filter(all_stats, region == do_region & ews == do_ews) %>% pull(value)
plot_it(dates, cases, ews, lab = "ews", title = do_ews)
}
Post-outbreak EWS
As the plots above show, most EWS get swamped out by the “memory” of outbreaks. To try and get around this, I am going to look at EWS after resetting them following outbreaks. I don’t think this is too contrived because real-world early detection systems would have to do the same thing.
How to ID outbreaks
Following Toby’s suggestion, I am going to look at the distribution of outbreak sizes (i.e., cumulative cases in each year). I can then fit a mixture distribution of two weighted exponentials to the data, with one exponential explaining the large outbreaks and one the small outbreaks. This provides an objective way to define outbreaks for EWS detection.
pop_data <- tibble(
region = c("Agadez (City)", "Maradi (City)", "Niamey (City)", "Zinder (City)"),
pop = c(118224, 267249, 1027000, 322935)
)
cusum_data <- measles_data %>%
group_by(region, year) %>%
summarise(cases = sum(cases)) %>%
left_join(pop_data)Joining, by = "region"ggplot(cusum_data, aes(x = cases/pop)) +
geom_histogram(aes(fill = region))
cusum_cases <- sort(cusum_data$cases / cusum_data$pop)
fit1 <- MASS::fitdistr(cusum_cases, "exponential")
x <- cusum_cases
mixexplik <- function(p,lambda1,lambda2) {
z <- p*dexp(x,lambda1) + (1-p)*dexp(x,lambda2)
return(-sum(log(z)))
}
mle_fit <- suppressWarnings( # ignore warnings about bad guesses
stats4::mle(minuslogl = mixexplik,
start= list(p = 0.2,
lambda1 = as.numeric(fit1$estimate),
lambda2 = as.numeric(fit1$estimate)),
method="Nelder-Mead")
)
get_prob <- function(x, rate1, rate2, p, x_scale, bin_width = 1){
p1 <- pexp((x+bin_width)/x_scale, rate1) - pexp((x-bin_width)/x_scale, rate1)
p2 <- pexp((x+bin_width)/x_scale, rate2) - pexp((x-bin_width)/x_scale, rate2)
mixp <- p*p1+(1-p)*p2
return(c(p1,p2,mixp))
}
x_scale <- 1e05
xstar <- round(x*x_scale)
probs <- sapply(xstar, FUN = get_prob, rate1 = mle_fit@coef["lambda1"],
rate2 = mle_fit@coef["lambda2"], p = mle_fit@coef["p"],
x_scale = x_scale, bin_width = 1)
bound <- qexp(p = 0.5, rate = mle_fit@coef["p"]*mle_fit@coef["lambda1"])
probs_df <- t(log10(probs)) %>%
as_tibble() %>%
dplyr::rename(
"log10prob_large" = V2,
"log10prob_small" = V1,
"log10prob_mix" = p
) %>%
dplyr::mutate(
incidence = x,
outbreak_size = ifelse(incidence < bound, "small", "large"),
plot_prob = ifelse(outbreak_size == "small", log10prob_small, log10prob_large)
)
ggplot(probs_df, aes(x = incidence)) +
geom_line(aes(y = log10prob_mix, color = "forestgreen")) +
geom_point(aes(y = plot_prob, fill = outbreak_size), size = 3, shape = 21, color = "white") +
geom_vline(aes(xintercept = bound), linetype = 2, color = "grey35") +
labs(x = "Case incidence (reports/person)",
y = expression(paste(Log[10]," probability"))) +
scale_fill_discrete(name = NULL, labels = c("Large outbreaks", "Small outbreaks")) +
scale_color_manual(name = NULL, labels = "GMM fit", values = "forestgreen") +
theme_minimal(base_size = 12)
OK, so now I can identify outbreak years in the untransformed (just cases) data for each city.
outbreak_ids <- cusum_data %>%
dplyr::mutate(
incidence = cases/pop,
outbreak_year = ifelse(incidence < bound, FALSE, TRUE)
) %>%
dplyr::select(region, year, outbreak_year)
measles_data <- measles_data %>%
left_join(outbreak_ids)Joining, by = c("region", "year")ggplot(measles_data, aes(x = date, y = cases, color = outbreak_year, group = year)) +
geom_line() +
facet_wrap(~region, scales = "free")
Let’s look at a test case of Niamey post 1995 leading up to the outbreak in 1999.
niamey_test <- measles_data %>%
filter(region == "Niamey (City)") %>%
filter(year > 1998 & year < 2000)
cases <- niamey_test$cases
variance <- spaero::get_stats(
x = cases,
center_trend = "local_constant",
center_kernel = "uniform",
center_bandwidth = 16,
stat_trend = "local_constant",
stat_kernel = "uniform",
stat_bandwidth = 16,
lag = 1,
backward_only = TRUE
)$stats$variance
plot_it(dates = niamey_test$date, cases = cases, ews = log(variance), lab = "log(variance)", title = NULL)
niamey_test <- measles_data %>%
filter(region == "Niamey (City)")
out_vars <- tibble()
for(do_year in unique(niamey_test$year)){
tmp_cases <- niamey_test %>%
filter(year == do_year) %>%
pull(cases)
tmp_stats <- spaero::get_stats(
x = tmp_cases,
center_trend = "local_constant",
center_kernel = "uniform",
center_bandwidth = 16,
stat_trend = "local_constant",
stat_kernel = "uniform",
stat_bandwidth = 16,
lag = 1,
backward_only = TRUE
)$stats
tmp_var <- tibble(
year = do_year,
date = filter(niamey_test, year == do_year)$date,
variance = tmp_stats$variance,
outbreak_year = filter(niamey_test, year == do_year)$outbreak_year
)
out_vars <- bind_rows(out_vars, tmp_var)
}
max_case_week <- niamey_test %>%
group_by(year) %>%
filter(cases == max(cases)) %>%
filter(week_of_year == min(week_of_year))
ggplot() +
geom_line(data = out_vars, aes(x = lubridate::week(date), y = log(variance),
group = year, color = outbreak_year)) +
geom_vline(data = max_case_week, aes(xintercept = week_of_year, color = outbreak_year), linetype = 2)
Hmmm, this seems promising. Perhaps I can fit a GAM through the trajectories and use model selection to determine if an interaction model (outbreak year*time interaction effect) fits better than a GAM with no interaction. Then can use Gavin’s methods for determining time points that are different. Model-based minimum time of difference would be the on-average indicator time, which could be compared to max cases within the year. Should do this with data (like above) and the simulations.
gam_data <- out_vars %>%
dplyr::select(date, variance, outbreak_year) %>%
dplyr::mutate(
week_of_year = lubridate::week(date),
log_variance = log(variance+0.00001),
outbreak_year = as.factor(outbreak_year)
) %>%
dplyr::select(-date)
test_formula <- as.formula(
log_variance ~ outbreak_year + s(week_of_year, by = outbreak_year)
)
gam_model <- mgcv::gam(formula = test_formula, data = gam_data)
pred_df <- tibble(
week_of_year = rep(1:52, times = 2),
outbreak_year = rep(as.factor(c(TRUE, FALSE)), each = 52)
)
preds <- predict(gam_model, newdata = pred_df, type = "response", se.fit = TRUE)
ests <- as.numeric(preds[[1]])
ses <- as.numeric(preds[[2]])
pred_df <- pred_df %>%
mutate(preds = ests,
se = ses)
ggplot(pred_df, aes(x = week_of_year, y = preds, color = outbreak_year)) +
geom_line() +
geom_line(aes(y = preds+(se*2)), linetype = 2) +
geom_line(aes(y = preds-(se*2)), linetype = 2)
But first, look at just case trajectories
ggplot(niamey_test, aes(x = week_of_year, y = cases, color = outbreak_year, group = year)) +
geom_line()
gam_data <- niamey_test %>%
dplyr::select(date, cases, outbreak_year) %>%
dplyr::mutate(
week_of_year = lubridate::week(date),
log_cases = log(cases+0.0001),
outbreak_year = as.factor(outbreak_year)
) %>%
dplyr::select(-date)
test_formula <- as.formula(
log_cases ~ outbreak_year + s(week_of_year, by = outbreak_year)
)
gam_model <- mgcv::gam(formula = test_formula, data = gam_data)
pred_df <- tibble(
week_of_year = rep(1:52, times = 2),
outbreak_year = rep(as.factor(c(TRUE, FALSE)), each = 52)
)
preds <- predict(gam_model, newdata = pred_df, type = "response", se.fit = TRUE)
ests <- as.numeric(preds[[1]])
ses <- as.numeric(preds[[2]])
pred_df <- pred_df %>%
mutate(preds = ests,
se = ses)
ggplot(pred_df, aes(x = week_of_year, y = preds, color = outbreak_year)) +
geom_line() +
geom_line(aes(y = preds+(se*2)), linetype = 2) +
geom_line(aes(y = preds-(se*2)), linetype = 2)
Extra
Maybe there is a variance value that serves as an indicator? Look at high AUCs from the simulations to see what the variances are.
ews <- read_csv("../../results/ews-emergence.csv") %>%
filter(city == "Maradi" & metric == "variance" & half == "second")Parsed with column specification:
cols(
metric = col_character(),
sim = col_integer(),
value = col_double(),
half = col_character(),
city = col_character(),
susc_discount = col_double()
)
|======== | 6%
|========= | 6%
|========= | 7% 1 MB
|========== | 8% 1 MB
|=========== | 8% 1 MB
|============ | 9% 1 MB
|============= | 10% 1 MB
|============== | 11% 1 MB
|=============== | 11% 1 MB
|=============== | 12% 1 MB
|================ | 13% 1 MB
|================= | 13% 1 MB
|================== | 14% 1 MB
|=================== | 15% 2 MB
|==================== | 15% 2 MB
|===================== | 16% 2 MB
|====================== | 17% 2 MB
|====================== | 17% 2 MB
|======================= | 18% 2 MB
|======================== | 19% 2 MB
|========================= | 20% 2 MB
|========================== | 20% 2 MB
|=========================== | 21% 2 MB
|============================ | 22% 2 MB
|============================= | 22% 3 MB
|============================== | 23% 3 MB
|=============================== | 24% 3 MB
|================================ | 25% 3 MB
|================================ | 25% 3 MB
|================================= | 26% 3 MB
|================================== | 27% 3 MB
|=================================== | 27% 3 MB
|==================================== | 28% 3 MB
|===================================== | 29% 3 MB
|====================================== | 29% 4 MB
|======================================= | 30% 4 MB
|======================================= | 31% 4 MB
|======================================== | 31% 4 MB
|========================================= | 32% 4 MB
|========================================== | 33% 4 MB
|=========================================== | 34% 4 MB
|============================================ | 34% 4 MB
|============================================= | 35% 4 MB
|============================================== | 36% 4 MB
|=============================================== | 36% 4 MB
|=============================================== | 37% 5 MB
|================================================ | 38% 5 MB
|================================================= | 38% 5 MB
|================================================== | 39% 5 MB
|=================================================== | 40% 5 MB
|==================================================== | 40% 5 MB
|===================================================== | 41% 5 MB
|====================================================== | 42% 5 MB
|======================================================= | 42% 5 MB
|======================================================= | 43% 5 MB
|======================================================== | 44% 5 MB
|========================================================= | 45% 6 MB
|========================================================== | 45% 6 MB
|=========================================================== | 46% 6 MB
|============================================================ | 47% 6 MB
|============================================================= | 47% 6 MB
|============================================================== | 48% 6 MB
|=============================================================== | 49% 6 MB
|================================================================ | 50% 6 MB
|================================================================ | 50% 6 MB
|================================================================= | 51% 6 MB
|================================================================== | 52% 7 MB
|=================================================================== | 52% 7 MB
|==================================================================== | 53% 7 MB
|===================================================================== | 54% 7 MB
|====================================================================== | 54% 7 MB
|======================================================================= | 55% 7 MB
|======================================================================= | 56% 7 MB
|======================================================================== | 56% 7 MB
|========================================================================= | 57% 7 MB
|========================================================================== | 58% 7 MB
|=========================================================================== | 58% 7 MB
|============================================================================ | 59% 8 MB
|============================================================================= | 60% 8 MB
|============================================================================== | 61% 8 MB
|=============================================================================== | 61% 8 MB
|=============================================================================== | 62% 8 MB
|================================================================================ | 63% 8 MB
|================================================================================= | 63% 8 MB
|================================================================================== | 64% 8 MB
|=================================================================================== | 65% 8 MB
|==================================================================================== | 65% 8 MB
|===================================================================================== | 66% 8 MB
|====================================================================================== | 67% 9 MB
|====================================================================================== | 67% 9 MB
|======================================================================================= | 68% 9 MB
|======================================================================================== | 69% 9 MB
|========================================================================================= | 70% 9 MB
|========================================================================================== | 70% 9 MB
|=========================================================================================== | 71% 9 MB
|============================================================================================ | 72% 9 MB
|============================================================================================= | 72% 9 MB
|============================================================================================== | 73% 9 MB
|=============================================================================================== | 74% 10 MB
|================================================================================================ | 75% 10 MB
|================================================================================================ | 75% 10 MB
|================================================================================================= | 76% 10 MB
|================================================================================================== | 77% 10 MB
|=================================================================================================== | 77% 10 MB
|==================================================================================================== | 78% 10 MB
|===================================================================================================== | 79% 10 MB
|====================================================================================================== | 79% 10 MB
|======================================================================================================= | 80% 10 MB
|======================================================================================================= | 81% 10 MB
|======================================================================================================== | 81% 11 MB
|========================================================================================================= | 82% 11 MB
|========================================================================================================== | 83% 11 MB
|=========================================================================================================== | 84% 11 MB
|============================================================================================================ | 84% 11 MB
|============================================================================================================= | 85% 11 MB
|============================================================================================================== | 86% 11 MB
|=============================================================================================================== | 86% 11 MB
|=============================================================================================================== | 87% 11 MB
|================================================================================================================ | 88% 11 MB
|================================================================================================================= | 88% 12 MB
|================================================================================================================== | 89% 12 MB
|=================================================================================================================== | 90% 12 MB
|==================================================================================================================== | 90% 12 MB
|===================================================================================================================== | 91% 12 MB
|====================================================================================================================== | 92% 12 MB
|====================================================================================================================== | 92% 12 MB
|======================================================================================================================= | 93% 12 MB
|======================================================================================================================== | 94% 12 MB
|========================================================================================================================= | 95% 12 MB
|========================================================================================================================== | 95% 12 MB
|=========================================================================================================================== | 96% 13 MB
|============================================================================================================================ | 97% 13 MB
|============================================================================================================================= | 97% 13 MB
|============================================================================================================================== | 98% 13 MB
|===============================================================================================================================| 99% 13 MB
|================================================================================================================================| 100% 13 MB
|================================================================================================================================| 100% 13 MBggplot(ews, aes(x = value)) +
geom_histogram() +
facet_wrap(~susc_discount, scales = "free")
med_var <- ews %>%
group_by(susc_discount) %>%
summarise(value = median(value))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